Preventing youths'high-risk sexual behaviors represent some of the most difficult and important challenges for social and behavioral scientists. This study will identify factors contributing to youths'high-risk sexual behaviors and their disparities. The objectives involve testing a causal model of sexual behaviors guided by the theory of triadic influence (TTI;Flay &Petraitis, 1994), a model well-tested in predicting other adolescent risk behaviors that specifies the operation of key proximal factors as well as more distal factors. Consistent with TTI, proximal factors will include self-efficacy, attitudes, and social norms;distal factors include global positive expectancies, depressive symptoms, parental relatedness, and academic orientation. Finally, potentially vital socio-cultural influences will be tested as ultimate (most distal) factors in the model. Global expectancies and cultural orientation in Latinos are two important factors of emphasis in the model that are understudied within comprehensive models of adolescents'sexual behaviors. The objectives will be accomplished through secondary analyses of Add Health data (National Longitudinal Study of Adolescent Health) collected from 1994-2002. These data (N = 20,745 at Wave I;N = 14,738 at Wave II;N = 15,197 at Wave III) afford very strong tests of the model from mid-adolescence through young adulthood. These data include reports from friends, partners and parents, and also have substantial proportions of Latino youth (including substantial Mexican Americans, but also over samples of Cuban American and Puerto Rican youth) and African Americans. Behavioral outcomes of focus at both follow-ups include reported early intercourse, unprotected sexual intercourse, and number of sexual partners. Additionally, reported unintended pregnancies, reported STDs, and positive tests on STD assays at Wave III will be included both to validate the behavioral reports and to provide further testing of the model's predictive efficacy. Cutting edge analysis techniques appropriate for longitudinal data such as structural equation, survival, and multi-level modeling will be employed to achieve the study objectives. By clarifying the causal paths leading to these behaviors and identifying which specific protective or risk factors are most critical for adolescent groups where disparities are apparent, the proposed project may aid in the development of more effective sexual-risk reduction programs designed for diverse adolescent populations. Adolescent sexual risk taking behaviors are major causes of premature death and disability in the US as well as lead to unplanned pregnancies. Despite evidence suggesting differing patterns of sexual behaviors between and within ethnic groups, the role of socio-cultural factors within comprehensive theory-guided models of the determinants of high-risk sexual behaviors have been relatively untested. By clarifying such factors like cultural orientation within such a theoretical model, this study will assist preventive programs in their efforts to address the most critical sexual risk taking deterrence factors for diverse adolescent groups.